Speech and Language Technology for Language Disorders

Speech and Language Technology for Language Disorders (2016) .. by Katharine Beals, etc




1 Overview of speech and language technologies
1.1 Introduction to speech and language technologies
1.2 Speech recognition
1.2.1 What is speech recognition?
1.2.2 Additional information – confidence and nbest
1.2.3 Types of language models Grammar-based language models Statistical language models
1.2.4 When can speech recognition help?
1.2.5 Limits of current technology
1.2.6 Availability of speech recognition technology
1.3 language understanding
1.3.1 What is natural language understanding?
1.3.2 Analyzing meaning
1.3.3 Information about intermediate structure
1.3.4 Converting language to action
1.3.5 Limits of current natural language understanding systems
1.3.6 Availability of natural-language-processing technology
1.4 Dialog systems
1.4.1 What are dialog systems?
1.4.2 When can dialog systems help?
1.4.3 Limitations of current systems
1.4.4 Availability of dialog system technologies
1.5 Text-to-speech
1.5.1 WhatisTTS?
1.5.2 Where could it help?
1.5.3 Limits of current systems
1.5.4 Availability of TTS systems
1.6 language generation
1.6.1 What is natural language generation?
1.6.2 Where could natural language generation help?
1.7 Text simplification
1.7.1 What is text simplification?
1.7.2 Where could text simplification help?
1.7.3 Limits of current systems
1.8 Complementary technologies
1.9 Conclusions

2 Overview of developmental language disorders

3 Technology for assessment and remediation of developmental language disorders
3.1 Linguistic technologies for assessing language needs
3.2 Linguistic technologies for remediation
3.2.1 Programs that address phonological processing
3.2.2 Programs that address comprehension, or receptive language
3.2.3 Programs that address productive language

4 Technology for task assessment, classroom accommodation, and communicative assistance of developmental language disorders
4.1 Linguistic technologies for task assessment – reading tasks in particular
4.2 Linguistic technologies for classroom accommodation
4.3 Assistive communication technologies for developmental language disorders

5 Conclusions and caveats about developmental language technology

6 Overview of acquired aphasia and disorders of word retrieval
6.1 Aphasia
6.1.1 Fluent aphasia
6.1.2 Non-fluent aphasia
6.1.3 Living with aphasia
6.2 Disorders of word retrieval in aphasia: “I know it but I cannot say it”
6.3 Approaches to treating word production disorders in aphasia
6.3.1 Lexical-semantic treatments
6.3.2 Lexical-phonological treatments
6.3.3 Semantic and phonological tasks are rarely pure

7 Software for aphasia: computer-assisted treatment of word retrieval deficits in aphasia
7.1 Technology for aphasia: what are the benefits?
7.2 Language software for aphasia: what is the evidence?

8 Software for aphasia: MossTalk Words® (MTW)
8.1 About MossTalk Words: a computer-implemented treatment
8.2 Research on MTW
8.3 Speech recognition in MTW-2
8.4 Conclusions
8.5 Commercial programs using speech recognition for word retrieval deficits in aphasia
8.6 The challenge
8.7 Moving beyond words
8.7.1 Speech-to-text/text-to-speech software
8.7.2 Role of the speech-language pathologist

9 Speech technology for aphasic sentence production disorders
9.1 Background: language production in non-fluent aphasia
9.1.1 Explanations Pathologically reduced short-term/working memory or resource diminution Weak activation of linguistic elements Difficulty with “thinking for speaking”
9.1.2 Approaches to treating sentence production in non-fluent aphasia Drill/practice exercises to increase activation of particular items or structures Treatments to improve thinking for speaking
9.2 Scope of the term “speech technology”
9.3 A tale of two programs
9.3.1 The “TS”: using speech technology for sentence production drills
9.3.2 SentenceShaper: enlarging the buffer for language How the program works A note about “lexical bootstrapping” SentenceShaper’s “aided effects”: theoretical implications Impact of narrative-based therapy with SentenceShaper Using SentenceShaper to train specific structures
9.3.3 Interleaving drill with narrative production: TS and SentenceShaper together
9.3.4 SentenceShaper research: some bottom lines
9.3.5 Future directions: using SentenceShaper to enhance life participation
9.4 Survey of speech technology for sentence production
9.4.1 Software that analyzes the user’s speech Goal: To give feedback about correctness and completeness Goal: To enable the user to engage in complex tasks Goal: To create a text transcript Goal: To analyze speech patterns for diagnostic purposes
9.4.2 Software that records and plays back the user’s speech Goal: To allow users to edit their speech Goal: To let users compare their speech to a model Goal: To enhance communication
9.4.3 Software that transmits the user’s speech
9.4.4 Helpful software not covered in this Chter Iconic communication aids Software for script training Speech recognition to support comprehension Software to track speech activity
9.5 Summary

10 Evaluating speech and language applications for language disorders
10.1 Use of the software
10.1.1 Some general considerations
10.1.2 Efficacy
10.1.3 Time to results (for remediation software)
10.1.4 Learning to use the software
10.1.5 User engagement
10.1.6 Responsiveness/robustness/implementation quality
10.1.7 Feedback
10.1.8 Accuracy of speech and language technologies
10.1.9 Usability in light of other issues
10.2 Contextual and support features
10.2.1 Cost, including initial cost and updates/new materials
10.2.2 Multiple users
10.2.3 Personalization and customization
10.2.4 Support/user community/documentation
10.2.5 Languages
10.2.6 Extensibility and growth
10.2.7 Record keeping
10.2.8 Assessment
10.2.9 Administration
10.2.10 Platform – is the product available on convenient, widely available platforms?
10.2.11 Evaluation strategy

11 Conclusions
11.1 Feedback
11.2 Assistive and remediative goals
11.3 Acquisition and repair
11.4 Reinforcements/rewards
11.5 Next steps

Authors’ biographies